Henry Ford Health Systems is involved in a variety of NIH research projects that require extremely computationally intensive tasks. These research projects fall into several categories: 1) Image synthesis, image processing, image patter recognition, and quantitative image analysis, this is particularly true in applications involving three- dimensional (spatial), four-dimensional (spatial and temporal) or multiple-dimensional (spatial, temporal, and multi-spectral or multi- parameter) data sets; 2) Modeling techniques utilizing three and four dimensional finite element calculations; and 3( Three dimensional image reconstruction. At the present time the only computer system available in the Henry Ford Health System with multi-processor capabilities and any computational power is a SUN 4/490 with a MVX accelerator. in the Image Analysis Laboratory. This system is interfaced to the hospital's Ethernet network which is a requirement for the transfer of images and the use of remote workstations and X-terminals. A major problem with this system is that the MVX is no longer supported by SUN. A hardware failure would jeopardize significant portions of the computational resources in the Image Analysis Laboratory. The objective of this proposal is to upgrade the research Image Analysis Laboratory at Henry Ford Hospital so that we can provide the computational power and speed, and data archiving that is required on large multi-dimensional data sets in an efficient manner so as to provide an effective research environment. The research Image Analysis Laboratory contains a local Ethernet network which is bridged to the Hospital's Ethernet which provides access to multiple users as well as image data sets from imaging devices and other non image research data sets. The upgrade will provide the necessary computational power and data archiving capabilities necessary for multiple investigations to carry out research objectives on multi-dimensional data sets, which otherwise they would have difficulty justifying on an individual basis.